Operator-based dynamic bandwidth management for teleoperation.

Abstract

Teleoperation has become a common feature in most systems and is now integrated into the control of collaborative robots’ swarms from a remote area. This thesis presents the design and development of a real time operator-based dynamic bandwidth management scheme for teleoperation of swarms of collaborative robots. This method is effective in complex teleoperation tasks that require the collaboration of several robots instead of a single robot. This thesis integrates personalization with bandwidth allocation through a novel two-level controller which dynamically updates the distribution of the available bandwidth among the robots to user (R2U), robot to robot (R2R) and user to robots (U2R) communication channels. This is accomplished through an optimization method using real-time observations. These observations reflect 1) changes in the personalizing factor (PF) through the changes in the level of disturbance encountered by the operator, 2) changes in quality of collaboration (QoC) of the executed task, along with 3) the occurrence of Interesting events (IEs) in the swarm's environment. This optimization ensures the effective accomplishment of the collaborative task with lower bandwidth consumption and better performance. The designed closed-loop two-level controller includes: Level 1, an event-based fuzzy logic controller triggered by the performance measures sent by the robot and results in the appropriate set of constraints; Level 2, a time-based triggered Lagrange optimization algorithm which determines the updated set of R2U, R2R and U2R communication rates. The designed operator-based dynamic bandwidth management algorithm is compared against the dynamic bandwidth management algorithm developed by Mansour et al. MATLAB simulations and experiments using humanoid robots were conducted. The comparison carried out in both validations demonstrates the superiority of the operator-based dynamic bandwidth management algorithm in terms of task and network performance. The performed simulations and experiments